clock menu more-arrow no yes mobile

Filed under:

The Debut of the DPI

Dylan Power Index? Defensive Paul Index? Deadly Panda Index? Might Need to Workshop It...

Villanova University vs University of Alabama, 2018 NCAA East Regional Playoffs Round 2 Set Number: X161800 TK1

What is the DPI?

The DPI is what I use to tell myself, “I got the game right, the players got it wrong,” it’s the ultimate fallback, the go-to ‘but actually’ when my bracket gets inevitably busted. In reality it’s 11 different numbers added up, the sum divided by 11, and the lower that final number is, the better.

The DPI is made up of 11 stats, compiled over a team’s season. The value each team gets is their respective rank across all of college basketball. I found these stats on or

  1. Scoring margin (how much a team wins/loses by)
  2. Rebound margin (how many total rebounds a team gets/how many their opposing team gets)
  3. Turnover margin (how many turnovers a team gives up/how many turnovers they force)
  4. Adjusted Offensive Efficiency (I don’t know what exactly goes into this, but whatever Purdue is good at)
  5. Adjusted Defensive Efficiency (The same as above except Texas Tech)
  6. Strength of Schedule (This is how you even out the mid-major numbers, or bump up teams in really good conferences, cough big 12)
  7. Offensive Effective Field Goal % (The quality of shots the offense gets)
  8. Defensive Effective Field Goal % (The quality of shots a defense gives up)
  9. Overall KenPom rank
  10. NET Ranking
  11. BPI Ranking

Now, a couple of these are intentionally redundant. The two that can get really crazy are rebound margin and turnover margin. Both are very important, it’s rare a bad rebounding team or a team that turns it over a lot makes a deep run. But I don’t want these two things to entirely skew the numbers.

For example, Marquette and Miami have the worst rebound margin in the field as a 9 and 10 seed. This one weakness shouldn’t turn them into worthlessness, but it should hurt their overall score.

Similarly, Creighton and Michigan State would have a terrible score since they rank so low in turnover margin. By adding a team KenPom rank, NET ranking and BPI ranking, those three scores are usually a close-knit number that pulls the average back to Earth. The BPI can get a little wonky (Houston, a 5 seed, is 2nd, and Colorado State, a 6 seed, is 81st.

So, after putting in 748 values by hand, dividing them by 11 to get 68 total numbers, this is what I found.

The ranked list from Baylor to Bryant
The Master Sheet (note: 2-pt% and 3-pt% were not calculated into the final average, but are good to know)

S-Tier Teams

To be in this tier, the key is to be good at everything and bad at nothing. Some of these teams are great at one thing, but most are just really solid at everything. These teams are dependable barring random variables like injuries, or just like the Basketball Gods hating you.


DPI Rank Seed Team DPI SCore
DPI Rank Seed Team DPI SCore
1 1 Baylor 12
2 5 Houston 17
3 4 UCLA 18
4 3 Tennesse 21
5 1 Gonzaga 22
6 3 Texas Tech 23
7 2 Kentucky 23
8 2 Auburn 23
9 2 Villanova 26
10 5 Saint Mary's 30
11 1 Arizona 30
12 1 Kansas 31

I can say with certainty that the 2022 champion is one of these 12 teams. Now there are some surprises here. Baylor at the top surprised me. They’re a little banged up, and the chance of repeating is extremely low, so I don’t think they’re a surefire bet to win the whole thing. Still very impressive what Scott Drew was able to do on the hangover of a championship in the best conference in basketball. (I don’t have them in my final four)

Let’s talk Houston. Their lowest value by far is their 96th rank in strength of schedule. Houston also has zero, yes zero, quad 1 wins. That’s super concerning. Not to mention they play one of the best double-digit seeds in the field in the first round. Take that 17 DPI score with a big dose of caution.

Other things that stood out, if you have a glaring weakness, you’re on the outside looking in. Purdue’s bad defense, Duke’s turnover margin, Wisconsin’s... everything? The lowest seeded teams in this section are two 5-seeds, Houston and Saint Mary’s. Saint Mary’s solid defense. A surprisingly deep WCC this year gave Saint Mary’s a higher SOS than some would expect, good for 56th in the country.

Proceed with Caution

These are the teams with some glaring problems that caused them to slide. Or even some things not in my original calculations that give me cause for concern.

Proceed with Caution

DPI Rank Seed Team DPI Score
DPI Rank Seed Team DPI Score
21 4 Illinois 47
24 3 Wisconsin 51
26 3 Purdue 53
38 6 Alabama 72
39 4 Providence 72

These are teams that I wouldn’t trust past the Sweet 16 and certainly not the Final Four. Illinois has an incredibly high turnover margin, and their first two games are against former Kansas chair-thrower Silvio DeSousa (now at Chattanooga) who could get Kofi Cockburn in foul trouble, and then the DPI darling Houston Cougars or Jelly Walker’s U** squad, a dangerous 12 seed. Illinois disappointed last year, expect similar results.

Purude STINKS at defense. Jaden Ivey is more of a combo guard on this team, and their second and third best players rarely share the floor together in Zach Edey and Trevion Williams. They have firepower, but so do other teams i.e., Iowa in the Big Ten Championship. Bad at defense is a bad recipe for March, just ask Fred Hoiberg.

Wisconsin and Providence can both be thrown into this barrel: luck. Not a bad thing to have this time of year, sometimes it is better to be lucky than good. Providence is the luckiest team in America according to KenPom and Wisconsin is 8th. I can’t break down everything that goes into these ratings, but essentially both of those teams won a lot more games than their numbers would indicate. Beating out close games that could’ve gone either way. You can ride their luck into the tournament or think that it runs out early. Both are quite a gamble.

Alabama has a bad scoring margin, a terrible turnover margin, and a bad defense. Not to mention they shoot a bunch of 3’s and miss a bunch of 3’s. One of the First Four teams always wins, this year it’ll be Notre Dame. Book it.

Cinderella Stories, or At Least Good for a Game

Whether it’s due to mid-major bias, or just poor seeding by the committee, these teams indicate that their numbers are better than what their seed is.


DPI Rank Seed Team DPI Score
DPI Rank Seed Team DPI Score
19 7 Murray State 46
20 10 Loyola Chicago 46
22 11 Virginia Tech 48
27 12 UAB 55
40 13 Vermont 74

Those top 3 teams have second week potential, it won’t be easy though. Murray State has to get through Kentucky, Loyola Chicago has to play Villanova, and Virginia Tech goes against a top ranked defense in Texas, to a top ranked offense in Purdue. But all three of these teams scored REALLY high.

Virginia Tech, conversely, to Wiscy and Providence, is one of the unluckiest teams in basketball even after a miraculous run to the ACC tournament championship, their luck could turn around in a hurry in the next few weeks. I think it’s better to side with ‘unlucky’ teams than the lucky ones. Virginia Tech has the numbers, just not the results, yet.

UAB has to face aforementioned Houston, but they are the best ranked 12 seed, seven spots higher than Indiana who will have to face a really solid Saint Mary’s.

The Elephant in the Room

So, I’m sure you’re thinking, ‘hey that’s great, but my bracket is already filled out, how does this make Iowa State look?’ Well, in a word: bad. Iowa State is tied as the second lowest ranked 11 seed with Notre Dame who need double OT to beat the lowest ranked 11 seed.

Their bad rebounding margin, paired with an abysmal offensive efficiency is holding the Cyclones top 10 defense and 14th ranked strength of schedule back. Even worse, LSU looks pretty solid. The DPI does not take into account head coaching vacancies or FBI investigations though.

Here’s a look at some teams ranked higher than the Cyclones: Chattanooga, Vermont, the already eliminated Wyoming Cowboys, and UAB is 18 spots ahead of us. sad exhale

Iowa State, sadly

DPI Rank Seed Team DPI Score
DPI Rank Seed Team DPI Score
45 11 Iowa State 82